EXAONE Path 2.0: Pathology Foundation Model with End-to-End Supervision

TL;DR

EXAONE Path 2.0 introduces a pathology foundation model trained with end-to-end supervision on whole-slide images, achieving state-of-the-art results in biomarker prediction with high data efficiency.

Contribution

It is the first to learn patch-level representations directly under slide-level supervision, surpassing self-supervised methods in data efficiency and performance.

Findings

Achieves state-of-the-art performance on 10 biomarker prediction tasks.

Uses only 37,000 WSIs for training, demonstrating high data efficiency.

Outperforms existing SSL-based approaches in pathology tasks.

Abstract

In digital pathology, whole-slide images (WSIs) are often difficult to handle due to their gigapixel scale, so most approaches train patch encoders via self-supervised learning (SSL) and then aggregate the patch-level embeddings via multiple instance learning (MIL) or slide encoders for downstream tasks. However, patch-level SSL may overlook complex domain-specific features that are essential for biomarker prediction, such as mutation status and molecular characteristics, as SSL methods rely only on basic augmentations selected for natural image domains on small patch-level area. Moreover, SSL methods remain less data efficient than fully supervised approaches, requiring extensive computational resources and datasets to achieve competitive performance. To address these limitations, we present EXAONE Path 2.0, a pathology foundation model that learns patch-level representations under direct slide-level supervision. Using only 37k WSIs for training, EXAONE Path 2.0 achieves state-of-the-art average performance across 10 biomarker prediction tasks, demonstrating remarkable data efficiency.